Intelligent Recognition Analysis of Chinese Herbal Medicine Images Using Deep Learning Algorithms

In view of the challenges faced by traditional manual recognition methods of Chinese medicinal materials, such as time and effort, easy to be subject to subjective interference, this study focused on the application of deep learning model to achieve high-precision and automatic recognition of Chinese medicinal materials images. This paper conducts experiments on a variety of deep learning models, including DarkNet, PPLCNet, ResNet, MobileNet, and ShuffleNet, comparing their performance on key metrics such as accuracy, recall, and F1 values. The results showed that DarkNet had the best performance in all indexes, especially in the processing of complex background and diversified Chinese herbal medicine images, and its recognition accuracy reached 97.4%. The research findings provide technical support for the further application of intelligent recognition of Chinese medicinal materials and lay the foundation for future deep learning applications in larger scale and more complex environments.